Run Social Media Research With 13 Agent Skills

A competitor teardown used to mean an afternoon of tab-hopping. This pack gives your agent the whole workflow — outlier posts, comment mining, transcript analysis, ad-library teardowns — against public social data. Each workflow ends in an artifact, not just raw JSON.

Why This Skill Matters

The README's skill table lists thirteen skills: outlier-post-finder, transcript-intelligence, comment-mining, competitor-social-research, ad-library-teardown, trend-discovery, influencer-prospecting, audience-research, social-listening-brief, product-demand-research, creator-profile-teardown, content-repurposing, and the scrapecreators-api data layer. Each targets one research job — the outlier finder judges posts against a creator's own baseline, the comment miner extracts objections and audience language, the ad-library skill reads active Meta, Google, and LinkedIn ads. Coverage spans TikTok, Instagram, YouTube, Reddit, X/Twitter, LinkedIn, Facebook, Threads, Bluesky, Pinterest, and Rumble, per the README.

Installation

npx skills add ScrapeCreators/social-media-research-skills

The README says this works with Claude Code, Cursor, OpenAI Codex, GitHub Copilot, Gemini CLI, Windsurf, VS Code, and other agents that support the Agent Skills spec. The skills need one API key:

export SCRAPECREATORS_API_KEY=sk_...

Get a key at scrapecreators.com — the skills are the workflow layer; ScrapeCreators is the scraping service underneath.

Real Workflow: Mine a Viral Reel's Comments

1. Install the pack and export your API key.
2. Ask, in the README's own example:

Mine the comments on this viral Instagram Reel. I want objections, questions, buying intent, and exact audience language.

3. The comment-mining skill returns a VOC report: themes, verbatim quotes, content ideas — the exact audience phrasing you would otherwise screenshot by hand.

A second pass with outlier-post-finder shows the pack's core idea, baseline-aware analysis: not just raw view counts — it flags posts that beat their creator's normal performance and extracts the repeatable pattern behind them.

Tips

• The twelve workflow skills use scrapecreators-api as their data layer when they need endpoint details — you describe the business outcome, not the API call.
• Design principle number three, cited outputs: reports include source URLs for posts, videos, ads, and comments whenever possible.
• Principle five, exact language: hooks, captions, and comment quotes are preserved verbatim, so voice-of-customer material stays quotable.
• Principle two, public-data only: don't ask for logged-in or private data — the README scopes these skills to public social data.

When Not to Use This

Every workflow runs through the ScrapeCreators API, so this is a dependency on an external service — if your research must run with zero outside services, these skills are not it. And by the README's own scoping, anything behind a login is out of bounds.

See the leaderboard for more skills.